1. Field of the Invention
The present invention generally relates to systems and methods for acquiring information about a defect on a specimen. Certain embodiments relate to a system that includes an optical subsystem configured to acquire topography information about the defect and an electron beam subsystem configured to acquire additional information about the defect.
2. Description of the Related Art
The following description and examples are not admitted to be prior art by virtue of their inclusion in this section.
Inspection processes are used at various times during a semiconductor manufacturing process to detect defects on a specimen such as a reticle and a wafer.
Inspection processes have always been an important part of fabricating semiconductor devices such as integrated circuits. However, as the dimensions of semiconductor devices decrease, inspection processes become even more important to the successful manufacture of acceptable semiconductor devices. For instance, as the dimensions of semiconductor devices decrease, detection of defects of decreasing size has become necessary since even relatively small defects may cause unwanted aberrations in the semiconductor devices. Accordingly, much work in the inspection field has been devoted to designing inspection systems that can detect defects having sizes that were previously negligible.
Another important part of yield control is determining the cause of the defects on the wafer or reticle such that the cause of the defects can be corrected to thereby reduce the number of defects on other wafers or reticles. Often, determining the cause of the defects involves identifying the defect type and other characteristics of the defects such as size, shape, composition, etc. Since inspection typically only involves detecting defects on the wafer or reticle and providing limited information about the defects such as location, number, and sometimes size, defect review is often used to determine more information about individual defects than that which can be determined from inspection results. For instance, a defect review tool may be used to revisit defects detected on a wafer or reticle and to examine the defects further in some manner either automatically or manually. Defect review can also be used to verify that defects detected by inspection are actual defects instead of, for example, noise and nuisance events.
Some examples of commonly used defect review tools include high resolution optical imaging systems, scanning electron microscopes (SEMS) and less commonly transmission electron microscopes. In order to be successful, the defect review tool must be able to accurately find the defects that are to be reviewed. For instance, if the defect review tool determines an incorrect location for a defect, during review at the incorrect position the defect review tool field of view (FOV) may be located on a non-defective portion of the specimen or a portion of a specimen containing a different defect. In this manner, when the defect review tool is erroneously positioned above a non-defective portion of the specimen, the defect review tool may determine that the detected defect is not an actual defect. Alternatively, if the defect review tool is erroneously positioned above a different defect on the specimen, the defect review results for the reviewed defect may be assigned to the wrong detected defect. As such, if the defect review tool cannot accurately find the defects that are to be reviewed, the defect review results may be substantially inaccurate and without user intervention (e.g., manually reviewing the defect review results) such inaccuracies may be largely undetected. Therefore, if the results of defect review are used to monitor and control semiconductor fabrication processes, the monitor and control may be ineffective or possibly even detrimental to the performance of the semiconductor fabrication processes.
Finding the defects in a defect review process is not trivial for a number of reasons. For instance, different coordinate sets are involved in determining the coordinates of the defect with respect to the FOV of the defect review tool. One set of coordinates is the set of coordinates in which the defect location is reported by the inspection tool that detected the defect. The defect location coordinates may be referenced to the inspection tool stage coordinates and translated such that the defect coordinates are referenced to the center of the specimen. An additional set of coordinates is the set of coordinates of the specimen with respect to the stage of the defect review tool. Since the specimen is typically moved from the inspection tool to the defect review tool, the coordinates of the specimen reported by the inspection tool will need to be translated to the review tool stage coordinates. During this coordinate mapping, there are systematic and random errors that need to be corrected before the review tool can locate the defect.
A number of different methods and systems have been developed to improve the defect finding step of defect review processes. To eliminate the systematic errors, normally wafer alignment, die corner registration, and defect coordinate deskew steps are performed. However, there are still random errors in the defect coordinates that are caused by the defect detection technology (such as technology for imaging or collecting scattered light from the defect) and defect location calculation and interpolation. In addition, different thermal conditions in the inspection tool during the defect detection and in the defect review tool during defect review may produce random error in the defect coordinates. One method to eliminate this random error that is commonly referred to as automatic defect locating (ADL) is performed using the defect review subsystem of the defect review tool. ADL may include generating a low resolution image (or large FOV image) of a specimen at approximately the location of a defect reported by the inspection tool using an optical microscope (OM) or a SEM of a defect review tool. The defect may then be redetected in the low resolution image. For instance, the large FOV image can be compared to a reference image to detect the defect in the low resolution image. A small FOV image of the defect may then be captured which is commonly used for defect classification purposes.
Such methods have a number of disadvantages. For instance, the FOV on the specimen during ADL is dependent on the defect coordinate inaccuracy of the inspection tool that inspected the specimen. In one such example, if the coordinates of the defect reported by the inspection tool are substantially inaccurate, then the FOV used for the low resolution imaging may need to be large enough to ensure that the defect is located in the image, but so large that defect redetection (particularly for relatively small defects) may not be possible or will take a relatively long time to perform. In addition, the sensitivity of ADL is limited by the parameters used for ADL including the FOV, pixel density, image integration time, noise in the image including specimen charging effects in SEM images, and the redetection algorithm that is used. Furthermore, ADL performed using SEM images will not be able to redetect defects that are inherently not visible to the SEM such as defects that are located below an upper surface of the specimen. Moreover, ADL performed using SEM images may also have a low redetection success rate for other types of defects such as low contrast defects, defects that are overwhelmed by unpredictable noise generated in the images, and defects that are too small to be redetected in low resolution SEM images.
An additional disadvantage of currently used methods and systems is that during ADL, imaging of the specimen by the SEM may cause contamination of each area that is imaged (e.g., both a defect die and a reference die). In addition, if all defect redetecting functions are performed in electron beam mode, then such repeated exposure of the specimen to the electron beam increases the potential for surface contamination and damage. The specimen must also be imaged at least twice by the SEM: at least once during ADL, and again during review. Therefore, such repeated exposure to the electron beam may increase the probability of damage to and contamination of the specimen. Furthermore, in order to mitigate the effects of ADL on the throughput of the defect review process, the low resolution imaging by the SEM may be performed with relatively high current and landing energy. Therefore, the portions of the specimen imaged during ADL may be subjected to conditions that are more likely to cause damage to and contamination of the specimen.
A different apparatus for reviewing defects on an object is illustrated in U.S. Pat. No. 6,407,373 to Dotan, which is incorporated by reference as if fully set forth herein. As described by Dotan, the apparatus includes a stage for receiving the object thereon, and both an OM and a SEM. The OM is used to redetect previously mapped defects on the object surface. Once the defect has been redetected, a translation system moves the stage a predetermined displacement such that the defect is positioned for review by the SEM.
Since the apparatus of Dotan does not use the SEM to redetect defects, the apparatus described by Dotan is less likely to cause electron damage and contamination of the specimen than the ADL methods described above. However, redetecting the defect as described by Dotan is disadvantageous for a number of reasons. For instance, such redetection may be quicker than the inspection process since only the positions in the defect map produced by inspection that indicate a potential defect are examined by the OM of the defect review apparatus. However, the defects still have to be redetected, which may limit the throughput, sensitivity, and accuracy of the defect location of the defect review process.
The apparatus disclosed by Dotan also has some of the disadvantages of ADL described above. For instance, the FOV on the specimen during the redetection described by Dotan is dependent on the defect coordinate inaccuracy of the inspection tool that inspected the specimen, which is disadvantageous as described further above. In addition, like ADL, the sensitivity of the defect redetection disclosed by Dotan is limited by the parameters of the OM including the FOV, the pixel density, and the redetection algorithm. Furthermore, most of the time, the imaging mode of the defect review tool is substantially different from the imaging mode used by the inspection tool to detect the defects. As a result of using this ADL method, there is no guarantee that the redetected defects in the FOV of the defect review tool are the same defects that are reported by the inspection tool. This error will cause false defect classification and incorrect defect root cause analysis.
Once a defect has been located in the defect review tool, information about the defect such as topography information (e.g., height) and non-topography information (e.g., width, shape, etc.) may be determined from data acquired by the SEM. Topography information about the defect can advantageously be used to determine whether the defect protrudes above the surface of the specimen or whether the defect extends into the specimen, which provides information about the defect type. A SEM can be used to determine such topography information by imaging the defect at an oblique angle of incidence or low angle illumination. Once the topography information has been acquired, the SEM can acquire additional, non-topography information about the defect with high angle illumination that is optically more favorable than low angle illumination for acquiring non-topography information. In particular, low angle illumination reduces the resolution of the data acquired by the SEM. Therefore, data responsive to non-topography information can be acquired by the SEM with higher resolution and greater accuracy using high angle illumination. Using the SEM in this manner to acquire both topography information and non-topography information about defects is disadvantageous for a number of reasons. For instance, using the two illumination modes to acquire the different types of information about the defects greatly reduces the throughput of the defect review process.
Accordingly, it may be advantageous to develop methods, defect review tools, and systems for locating a defect in a defect review process that are independent of the defect coordinate accuracy of an inspection tool, are not limited by the parameters of the imaging subsystem used to image the specimen during defect locating, do not cause charging and/or contamination of the specimen during defect locating, and do not reduce the throughput of the defect review process. It would also be advantageous to develops systems and methods for acquiring information about a defect on a specimen that can acquire topography information about the defect without using an electron beam technique thereby increasing the throughput of the methods and systems.